Hi,
So I'm implementing a pretty standard tuxedo pipeline on paired-end mouse data. I went along as follows.
Using the standard mouse build (37) from Ensemble and the associated annotation data I did the following.
cutadapt -> tophat2 -> cufflinks -> cuffmerge -> cuffquant -> cuffdiff -> cummerbund.
Tophat2 is giving me roughly 85% overall mapping, 5-10% multimap. Cufflinks was run on around 100 files and those were merged with cuffmerge. All of the bam files were then run with cuffquant using the gtf from cuffmerge.
To test out an initial dataset I just used two sample conditions (2 replicates each) to do cuffdiff. 8 files. (4 control, 4 corresponding experimental case).
Now, I load this up in cummeRbund and see the following:
So this is my first time working with mouse or cufflinks pipeline, but somehow these numbers don't feel right. So I checked it out and at least I found that mouse has only around 23,000 genes, so thats wrong for sure.
Could someone explain to me what sort of numbers I should be seeing here and why at least cuffdiff/cummRbund is showing approximately double the number of genes that should exist in the genome I'm looking at?
I appreciate any feedback anyone can provide.
So I'm implementing a pretty standard tuxedo pipeline on paired-end mouse data. I went along as follows.
Using the standard mouse build (37) from Ensemble and the associated annotation data I did the following.
cutadapt -> tophat2 -> cufflinks -> cuffmerge -> cuffquant -> cuffdiff -> cummerbund.
Tophat2 is giving me roughly 85% overall mapping, 5-10% multimap. Cufflinks was run on around 100 files and those were merged with cuffmerge. All of the bam files were then run with cuffquant using the gtf from cuffmerge.
To test out an initial dataset I just used two sample conditions (2 replicates each) to do cuffdiff. 8 files. (4 control, 4 corresponding experimental case).
Now, I load this up in cummeRbund and see the following:
CuffSet instance with:
2 samples
52615 genes
220501 isoforms
106928 TSS
49476 CDS
52615 promoters
106928 splicing
21223 relCDS
2 samples
52615 genes
220501 isoforms
106928 TSS
49476 CDS
52615 promoters
106928 splicing
21223 relCDS
Could someone explain to me what sort of numbers I should be seeing here and why at least cuffdiff/cummRbund is showing approximately double the number of genes that should exist in the genome I'm looking at?
I appreciate any feedback anyone can provide.